We have preliminary data in the mouse LP-BM5 model of AIDS infection (MAIDS) showing that AZT, the only FDA approved treatment of AIDS at this point, reduced virus titer in spleen and lymph nodes by 100-1000 fold, as well as serum immunoglobulin levels, B, CD4+, and Macl+ cell proliferation, and splenomegaly and lymphadenopathy. After six weeks of AZT therapy both the number of CD8+ cells were increased and survival time was extended by 5-6 weeks as compared to untreated virus infected control mice. This model has proven to be very reproducible, inexpensive, and lends itself to specific and sophisticated immunoanalysis. We propose to investigate several means of improving current in vivo anti-retroviral therapy using the LP-BM5 MAIDS pre-clinical model. We will test additional treatment schedules and doses of AZT for maximum effectiveness. Other antiviral drugs will be used both alternately and concomitantly with, AZT in an effort to improve efficacy. Selected BMR will be studied for their ability to enhance AZT action. We will extend our Stanford group's (Blaschke and Robinson) in vitro observation that lithium enhanced AZT activity against HIV in molt4 cells by testing a combination in AZT and LiCl in the MAIDS model. In addition, new drugs including nucleotide analogs, specific protease and glycosidase inhibitors from industrial sources will be assessed by pre-clinical testing in vitro and in vivo as they are available. We will also attempt to improve drug delivery and effectiveness via drug encapsulation in liposomes. The impact of murine CMV both as a latent and acute infection will be explored in the MAIDS model. This is particularly relevant because of the potential lethal sequelae of CMV infections in human AIDS patients. Information on optimum dose, route scheduling, the effectiveness of combination therapy and the impact of at least one other infectious agent in the model will be useful in designing more rational future clinical protocols.